> > I was reading a year+ old article where Steve Scott (CRAY CTO) was > > interviewed here: > > https://www.nextplatform.com/2016/01/08/cray-cto-connects-the-dots- > on-future-interconnects/ > > > > At a certain point Steve mentions (or at least that's what the > article > > claims): > > > > /InfiniBand has another limit in terms of the number of logical IDs, > or > > LIDs, it can support, which is around 48,000 end points, and for the > > most point, people have stayed within that limit. There is an > extended > > version that has some higher packet overhead that can go to higher > > scalability./ > > > > Is this true? Is there any reference that one can point me to for > the > > part of the "extended version that has some higher packet overhead > that > > can go to higher scalability"? > > > > To the best of my knowledge, there is no extended version defined in > the > > InfiniBand specification. What is mentioned is that /The unicast LID > > range is a flat identifier space defined as 0x0001 to 0xBFFF./ > > Right, there is no any extended version of that spec. > As far as I know, Intel's OmniPath technology is marketed as new and > extended version of Infiniband [1], so it is probably that he > referenced to it. At one point there was talk about using the GRH to extend IB subnet address space. I can't remember what became of that proposal, but I can't quickly find any mention of it in the spec. But adding the GRH would result in a small increase to packet overhead. - Sean -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
